Spaces:
Sleeping
Sleeping
| # AUTOGENERATED! DO NOT EDIT! File to edit: ../weed_classifier.ipynb. | |
| # %% auto 0 | |
| __all__ = ['learn', 'labels', 'article', 'title', 'description', 'examples', 'interpretation', 'enable_queue', 'predict'] | |
| # %% ../weed_classifier.ipynb 1 | |
| from fastai.vision.all import * | |
| import gradio as gr | |
| import skimage | |
| # %% ../weed_classifier.ipynb 2 | |
| learn = load_learner('export.pkl') | |
| # %% ../weed_classifier.ipynb 3 | |
| labels = learn.dls.vocab | |
| def predict(img): | |
| img = PILImage.create(img) | |
| pred,pred_idx,probs = learn.predict(img) | |
| return {labels[i]: float(probs[i]) for i in range(len(labels))} | |
| # %% ../weed_classifier.ipynb 5 | |
| article = """<div>This model was trained on the dataset contains 5,539 images of crop and weed seedlings. | |
| The images are grouped into 12 classes. | |
| These classes represent common plant species in Danish agriculture at different growth stages. | |
| A pretrained ResNet34 model was fine-tuned using fastai vision library. The error rate achived on the validation dataset is 2.5% after 4 epochs. | |
| Currently the model does not generalize well on the out-of-domain images. Probably because of the specificity of the dataset. | |
| The images in the dataset are of a single plant taken from the top and on a single background of a grit surface. | |
| <p style='text-align: center'><a href='https://tmabraham.github.io/blog/gradio_hf_spaces_tutorial' target='_blank'>Blog post</a></p> | |
| </div>""" | |
| # %% ../weed_classifier.ipynb 6 | |
| title = "Weed Classifier" | |
| description = "<h3 style='text-align: center'>A weed classifier trained on the Kaggle V2 Plant Seedling dataset with fastai.</h3>" | |
| examples = ['SugarBeet.png'] | |
| interpretation='default' | |
| enable_queue=True | |
| # %% ../weed_classifier.ipynb 7 | |
| gr.Interface(fn=predict,inputs=gr.inputs.Image(shape=(512, 512)),outputs=gr.outputs.Label(num_top_classes=3),title=title,description=description,article=article,examples=examples,interpretation=interpretation).launch(enable_queue=enable_queue) | |